{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/gradient-starvation-a-learning-proclivity-in","title":"Gradient Starvation: A Learning Proclivity in Neural Networks","arxiv_id":"2011.09468","date":"2020-11-18","proceeding":"NeurIPS 2021 12","authors":["Mohammad Pezeshki","Sékou-Oumar Kaba","Yoshua Bengio","Aaron Courville","Doina Precup","Guillaume Lajoie"],"abstract":"We identify and formalize a fundamental gradient descent phenomenon resulting in a learning proclivity in over-parameterized neural networks. Gradient Starvation arises when cross-entropy loss is minimized by capturing only a subset of features relevant for the task, despite the presence of other predictive features that fail to be discovered. This work provides a theoretical explanation for the emergence of such feature imbalance in neural networks. Using tools from Dynamical Systems theory, we identify simple properties of learning dynamics during gradient descent that lead to this imbalance, and prove that such a situation can be expected given certain statistical structure in training data. Based on our proposed formalism, we develop guarantees for a novel regularization method aimed at decoupling feature learning dynamics, improving accuracy and robustness in cases hindered by gradient starvation. We illustrate our findings with simple and real-world out-of-distribution (OOD) generalization experiments.","url_abs":"https://arxiv.org/abs/2011.09468v4","url_pdf":"https://arxiv.org/pdf/2011.09468v4.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"gradient-starvation-a-learning-proclivity-in","repo_url":"https://github.com/mohammadpz/Gradient_Starvation","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"gradient-starvation-a-learning-proclivity-in","repo_url":"https://github.com/facebookresearch/DomainBed","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"out-of-distribution-generalization","task_name":"Out-of-Distribution Generalization"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2011.09468","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2011.09468"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/facebookresearch/DomainBed","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/mohammadpz/Gradient_Starvation","reach":null}],"summary":{"ran":1,"unverified":1},"by_repo_kind":{"official":{"samples":1,"ran":1,"repositories":1},"listed":{"samples":1,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"1f43cc4e96ba9601","entry":"LossComputer","repo":"mohammadpz/Gradient_Starvation","repo_kind":"official","path":"Table_2/loss.py","file_url":"https://github.com/mohammadpz/Gradient_Starvation/blob/HEAD/Table_2/loss.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1f43cc4e96ba9601"}},{"code_sha256_prefix":"afcee187a7194f14","entry":"SD","repo":"facebookresearch/DomainBed","repo_kind":"listed","path":"domainbed/algorithms.py","file_url":"https://github.com/facebookresearch/DomainBed/blob/HEAD/domainbed/algorithms.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"afcee187a7194f14"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}